Attitudes, Persuasion, & Behavior

Stage of Processing: From a Social Message to Behavior


One of my first lines of research examined the sequence of cognitive and motivational events that mediate the impact of persuasive communication on attitudes and behavior. One set of studies (Albarracin & Wyer, 2001) demonstrated that when people have time to think about a persuasive behavior-related message, they form beliefs and evaluations about the behavior outcomes described. Then, these cognitions combine to influence attitudes and behavior. In contrast, when people are not able to think carefully about the message content, they base their attitudes and behavior on the affective feelings they are experiencing at the time. That is, they attribute these feelings to their opinion on the advocated position. In this case, their beliefs and evaluations of specific arguments are formed subsequently to justify the attitude they have already formed. In both cases, people form attitude-consistent beliefs. In one case, however, their beliefs are the determinants of their attitude; in the other case, their beliefs are the consequences.


To my knowledge, this research was the first to detail the cognitive processes that mediate responses to a persuasive message on attitudes, beliefs, and behavior. In particular, it was the first to reveal the order in which these responses occur. This demonstration was important because other models of persuasion assume that beliefs are not involved when people have lower cognitive capacity. In contrast, my work showed that people rationalize their attitudes and behaviors by forming consistent beliefs. These beliefs are likely to strengthen the effects of affective reactions on attitudes. As a result, the influence of affect may be just as long-lasting as the influence of the communicated arguments, if not more. This conclusion is again contrary to the dual-process assumption that heuristic processing has short-lived effects on attitudes. More recent work in the area of persuasion has identified comparative processes leading to attitude change (Albarracin, Wallace, & Hart, 2012), the role of motivational processes in attitude change (Albarracin et al., 2011), as well as belief rationalization in greater detail (Albarracin & McNatt, 2005).


There are other reasons why initial impact on attitudes rather than beliefs can lead to greater maintenance of attitude and behavior change. In another study (Albarracin & McNatt, 2004), we used our previously developed method to study the effects of past behavior (Albarracin & Wyer, 2000). This method excludes the potential cognitive activity that takes place at the time of the behavior. Specifically, participants were led to believe that they had unconsciously supported or opposed a social policy. This feedback had direct effects on attitudes about the policy and its expected outcomes. Interestingly, the direct effects on attitudes were relatively global and easy to recall. As a result, self-perception effects lasted longer than more specific thoughts about the outcomes of the policy. This finding also departed from prior hypotheses  stating that elaboration necessarily produces maintenance of attitude change.


Another series of studies considered the way in which affect contributes to the cognitive response to a persuasive message (Albarracin & Kumkale, 2004). The use of affect as information is well established (Schwarz & Clore, 1983). However, we analyzed the sequence of cognitive activities that underlie this influence. Specifically, we distinguished between the identification of one’s experienced affect and the evaluated relevance of this affect to a particular judgment. We experimentally varied participants’ information processing capacity and motivation to obtain various levels of processing. We showed that the extraneous affect that participants experience has little impact when their level of processing is high and they recognize that affect is not relevant. Extraneous affect also has little impact when level of processing is low and people are unable to attend to this affect. Thus, level of processing has a non-monotonic influence on the impact of affect: impact is greater when processing level is moderate (when people are able to identify affect but are unable to evaluate its relevance) than when processing level is high or low. These findings extend our understanding of the conditions in which affect impacts judgments.